language:
- en
license: apache-2.0
library_name: atommic
datasets:
- CC359
thumbnail: null
tags:
- image-reconstruction
- KIKINet
- ATOMMIC
- pytorch
model-index:
- name: REC_KIKINet_CC359_12_channel_poisson2d_5x_10x_NNEstimationCSM
results: []
Model Overview
KIKINet for 5x & 10x accelerated MRI Reconstruction on the CC359 dataset.
ATOMMIC: Training
To train, fine-tune, or test the model you will need to install ATOMMIC. We recommend you install it after you've installed latest Pytorch version.
pip install atommic['all']
How to Use this Model
The model is available for use in ATOMMIC, and can be used as a pre-trained checkpoint for inference or for fine-tuning on another dataset.
Corresponding configuration YAML files can be found here.
Automatically instantiate the model
pretrained: true
checkpoint: https://huggingface.co/wdika/REC_KIKINet_CC359_12_channel_poisson2d_5x_10x_NNEstimationCSM/blob/main/REC_KIKINet_CC359_12_channel_poisson2d_5x_10x_NNEstimationCSM.atommic
mode: test
Usage
You need to download the CC359 dataset to effectively use this model. Check the CC359 page for more information.
Model Architecture
model:
model_name: KIKINet
num_iter: 2
kspace_model_architecture: UNET
kspace_in_channels: 2
kspace_out_channels: 2
kspace_unet_num_filters: 16
kspace_unet_num_pool_layers: 2
kspace_unet_dropout_probability: 0.0
kspace_unet_padding_size: 11
kspace_unet_normalize: true
imspace_model_architecture: UNET
imspace_in_channels: 2
imspace_unet_num_filters: 16
imspace_unet_num_pool_layers: 2
imspace_unet_dropout_probability: 0.0
imspace_unet_padding_size: 11
imspace_unet_normalize: true
dimensionality: 2
reconstruction_loss:
l1: 0.1
ssim: 0.9
estimate_coil_sensitivity_maps_with_nn: true
Training
optim:
name: adamw
lr: 1e-4
betas:
- 0.9
- 0.999
weight_decay: 0.0
sched:
name: CosineAnnealing
min_lr: 0.0
last_epoch: -1
warmup_ratio: 0.1
trainer:
strategy: ddp_find_unused_parameters_false
accelerator: gpu
devices: 1
num_nodes: 1
max_epochs: 20
precision: 16-mixed
enable_checkpointing: false
logger: false
log_every_n_steps: 50
check_val_every_n_epoch: -1
max_steps: -1
Performance
To compute the targets using the raw k-space and the chosen coil combination method, accompanied with the chosen coil sensitivity maps estimation method, you can use targets configuration files.
Evaluation can be performed using the evaluation script for the reconstruction task, with --evaluation_type per_slice.
Results
Evaluation against RSS targets
5x: MSE = 0.003224 +/- 0.003526 NMSE = 0.04931 +/- 0.05484 PSNR = 25.43 +/- 4.157 SSIM = 0.7882 +/- 0.08686
10x: MSE = 0.004036 +/- 0.0038 NMSE = 0.06195 +/- 0.06049 PSNR = 24.37 +/- 3.88 SSIM = 0.7419 +/- 0.1053
Limitations
This model was trained on the CC359 using a UNet coil sensitivity maps estimation and might differ from the results reported on the challenge leaderboard.
References
[1] ATOMMIC
[2] Beauferris, Y., Teuwen, J., Karkalousos, D., Moriakov, N., Caan, M., Yiasemis, G., Rodrigues, L., Lopes, A., Pedrini, H., Rittner, L., Dannecker, M., Studenyak, V., Gröger, F., Vyas, D., Faghih-Roohi, S., Kumar Jethi, A., Chandra Raju, J., Sivaprakasam, M., Lasby, M., … Souza, R. (2022). Multi-Coil MRI Reconstruction Challenge—Assessing Brain MRI Reconstruction Models and Their Generalizability to Varying Coil Configurations. Frontiers in Neuroscience, 16. https://doi.org/10.3389/fnins.2022.919186